OpenAI's new safety hire says losing control of AI would be 'catastrophic' and that 'most people could die'

OpenAI has appointed its first safety researcher, Paul Christiano, who issued a stark warning about the dangers of AI. Christiano believes that if humans lose control of AI, most people could die. He emphasized the need for global coordination to tackle this risk. Christiano, who previously worked at OpenAI and the federal government, stated that the company and the industry are not on track to mitigate this risk to an acceptable level.

He highlighted the rapid acceleration in AI capabilities and the ongoing challenges in alignment as factors contributing to this imminent threat. He warned that despite recent progress in AI research, there is a meaningful risk of a catastrophic and irreversible loss of control in the near term. Christiano called on all frontier labs to improve their safety oversight and stressed the importance of worldwide collaboration to address the risks posed by AI.